研究目的
To model ?uctuation process characteristics of photovoltaic (PV) outputs using a novel mixed Gaussian model with the expectation maximization (EM) algorithm.
研究成果
The TM-G model can be used to fit the fluctuation of PV output accurately, providing a more accurate fitting effect than single probability distribution. This work offers a universal methodology for analyzing fluctuation characteristics of PV outputs, useful for wide-area distributed PV aggregation analysis.
研究不足
The study focuses on the fluctuation characteristics of PV outputs in specific conditions and locations, which may not be universally applicable. The complexity of the model increases with the number of mixtures, potentially affecting training time.
1：Experimental Design and Method Selection:
The study employs a mixed Gaussian model with the EM algorithm to analyze the fluctuation characteristics of PV outputs. The theoretical output part and a volatility output part are considered based on available measurements.
2：Sample Selection and Data Sources:
Historical data of Xichang City (1 d, 15 min/point) on July 20th, 21st, 25th, and August 9th, 2018, are used for simulation.
3：List of Experimental Equipment and Materials:
MATLAB is used as the simulation experiment platform.
4：Experimental Procedures and Operational Workflow:
The difference between the measured data of PV output and its theoretical outputs is computed to obtain random components. The EM algorithm is then used to determine the weight of different Gaussian distribution functions, and the mixed Gaussian model is obtained by linearly superimposing these Gaussian functions with the weight.
5：Data Analysis Methods:
The effectiveness of the proposed model is verified based on simulation results, comparing it with other traditional models including t location-scale (TLS) distribution model.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
